Generate n8n Workflows with DeepWiki and Copilot

Recently, I came across an intriguing post from an influencer demonstrating how to generate n8n workflows using Claude. This method involved creating and saving various files within Claude projects.

While the idea of using AI to generate n8n workflows is certainly appealing, this particular approach has a significant drawback: the files used are based on the living code of the n8n repository. With over 10,000 files, you can imagine how frequently the code changes, even if not every part updates constantly. Relying on static files from a rapidly evolving codebase can quickly lead to outdated or broken workflows.

It's worth noting that n8n actually offers its own AI-powered workflow generator. However, my last experience with it wasn't great; it felt like a 'black box' and didn't perform as expected.

Exploring a Different Way: AI Tools with Continuous GitHub Access

Given these challenges, I became curious if we could leverage other AI tools that already have continuous access to the latest code on GitHub. Could they help us with the same goal: generating a workflow for a specific use case?

For instance, a use case might be: "Read topics from an Excel sheet, conduct research on each topic (in a loop), summarize them in 300 words, and then post the summary on LinkedIn."

Two very popular tools I've mentioned in previous blog posts that fit this description are DeepWiki (from Devin) and GitHub Copilot.